--- title: "Meituan executives interpret Q4 financial report: will continue to firmly promote the research and development of self-developed large language model LongCat" type: "News" locale: "en" url: "https://longbridge.com/en/news/280639096.md" description: "Meituan released its fourth-quarter financial report for 2025, with annual revenue of 364.9 billion yuan, a year-on-year increase of 8%, but due to intensified competition, it reported a net loss of 23.4 billion yuan. CEO Wang Xing emphasized during the analyst conference call that artificial intelligence will bring disruptive changes, and Meituan will actively invest in AI research and development, especially in its self-developed large language model LongCat, to enhance its core business capabilities" datetime: "2026-03-26T13:20:39.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/280639096.md) - [en](https://longbridge.com/en/news/280639096.md) - [zh-HK](https://longbridge.com/zh-HK/news/280639096.md) --- > Supported Languages: [简体中文](https://longbridge.com/zh-CN/news/280639096.md) | [繁體中文](https://longbridge.com/zh-HK/news/280639096.md) # Meituan executives interpret Q4 financial report: will continue to firmly promote the research and development of self-developed large language model LongCat Meituan released its performance report for the fourth quarter and the full year of 2025. The announcement showed that Meituan achieved an annual revenue of 364.9 billion yuan, a year-on-year increase of 8%. Affected by the "involution" competition in the instant retail industry, Meituan's performance turned from profit to loss, with a net loss of 23.4 billion yuan for the year and an operating loss of 17 billion yuan. After the financial report was released, Meituan's Chairman and CEO Wang Xing, and Senior Vice President and CFO Chen Shaohui participated in the subsequent analyst conference call to interpret the financial report. **The following is a transcript of the Q&A session from the conference call:** **Goldman Sachs Analyst Ronald Keung:** I would like to ask management about the competitive landscape of artificial intelligence. With the arrival of a new wave of technology, we see intensified competition in the industry around entry points and traffic. How does the company view this trend and its future development? Additionally, from a risk perspective, could this competition potentially weaken the company's position as a major traffic entry point in the application era? What strategies or plans does management have in place to address this potential risk? Furthermore, could management provide an update on Meituan's latest progress in AI intelligence and large models? **Wang Xing:** In previous quarterly earnings calls, I have clearly stated that I believe artificial intelligence is likely to bring disruptive changes across various fields. In this AI revolution, the only reasonable and appropriate strategy is "proactive offense," rather than "passive defense." However, this does not mean we will rush to become what is called a "token factory." On the contrary, we view AI technology as a strategic opportunity that can help us enhance, strengthen, and even reshape Meituan's product capabilities in the core business of local life services. Today, I would like to take this opportunity to share my thoughts in detail. First, AI requires significant investment. Since the beginning of 2023, we have made substantial investments in capital expenditures and AI talent, most of which are used to build self-developed models. Among Chinese companies, apart from those in cloud computing, Meituan can be said to be one of the leading companies in terms of investment in the AI field. In fact, our investment in this direction has continued for more than three years, which has undoubtedly had some impact on the company's balance sheet and cash flow. However, we will continue to firmly advance the research and development of our self-developed large language model LongCat. We believe that only by having the capability to develop our own models can Meituan understand the real world more accurately based on massive data. At the same time, we are also collaborating with third-party models. We are working hard to upgrade the Meituan App into an AI-driven application to better meet users' end-to-end needs in local life and instant retail. In our view, the so-called competition for "super gateways" in the industry essentially revolves around whether the platform can accurately understand user needs and efficiently execute tasks. However, this issue is far more complex than simply creating a "smart chatbot." The user scenarios in the local life service industry are extremely complex, involving a large amount of recommendation information and massive real-time data from small and medium-sized merchants The digitalization level of these merchants is very limited. The reality is that the data on the merchant side and even some user side has not yet been fully digitized. If we want everyone to truly benefit from AI, the prerequisite is to complete the digitization of the aforementioned data, which is precisely the work we have been continuously promoting for many years. A large number of merchants are already operating on our digital system, so it can be said that Meituan has unparalleled data acquisition capabilities. In addition, as a mobile local life service platform, we must also deeply participate in the management of fulfillment and delivery services, and cannot just stop at creating chatbots. In this regard, I believe that general AI currently cannot reliably manage and ensure service experiences in the real world. Meituan has been built around the real world from the very beginning. We have accumulated a large amount of relevant data, including merchant POI (Point of Interest) data, dynamic and real-time merchant operation data, as well as the most comprehensive and authentic user review data, etc. I believe that our deep accumulation in delivery networks, instant delivery systems, business expansion and operational capabilities, retail supply chains, and future intelligent technologies such as drones and unmanned vehicles will provide significant advantages in bridging the connection between AI and the real world. Moreover, our self-developed models are gradually catching up with open-source models, and the capabilities of AI entities are rapidly evolving, which will help us achieve seamless integration of information between the digital world and the real world. For example, we recently launched the AI assistant "Xiao Tuan" to all users in our app. Prior to this, we also released a standalone AI application "Xiao Mei." With the launch of "Xiao Tuan," hundreds of millions of Meituan App users will benefit from its AI capabilities. "Xiao Tuan" covers all local life service categories on the platform, allowing users to express their needs in a more natural way. In the past, users might have been more accustomed to searching through keywords; now, with enhanced AI capabilities, "Xiao Tuan" can understand users' longer and more natural language inputs and can fully utilize all data resources within the Meituan App. Let me give a specific example. You can also imagine that this scenario is actually very common: users often search for restaurants through the Meituan App, but in many cases, we need to understand users' actual needs more deeply. For instance, I am currently working in Wangjing, northeast of Beijing, while my friend is working in Zhongguancun, to the west of Beijing. We plan to have lunch together, but we only have two hours in total. So I might express a need like this: can you help us find a restaurant located between Zhongguancun and Wangjing, specializing in Sichuan cuisine, with a spicy flavor, and also convenient for parking? This is actually a very real and natural user need, but in the past, under the old keyword-based search model, users could not pose such complex questions to the platform; now, with enhanced AI capabilities, such needs can be understood and processed by the platform. However, to truly answer such questions, merely having a "smart chatbot" is far from enough. The system needs to have multiple capabilities: first, we need to understand user intent; second, we need to have map information, POI data, and traffic information; At the same time, we cannot only know "what restaurants are available," but we must also understand the specific conditions of the restaurants, such as the characteristics of the dishes and real-time reception capacity. Otherwise, the system may recommend a very popular restaurant, but the user may not be able to book a seat or a private room, which clearly does not meet the user's real needs. Therefore, to truly satisfy user needs, our system must be able to access and integrate a large amount of information from the real world. AI is helping us connect this data to provide users with a better experience. Relying on Meituan's comprehensive and authentic merchant database, "XiaoTuan" can now not only answer basic questions such as merchant location and business hours but also provide more information such as store facilities. At the same time, "XiaoTuan" can utilize real user reviews and recommendations to provide users with more valuable decision-making references. With this enhanced reasoning capability, "XiaoTuan" can handle more personalized user needs and generate one-stop solutions for dining, entertainment, travel, and more. We are also pleased to see that during the Spring Festival, "XiaoTuan" has been able to effectively respond to user needs. Of course, this is just the beginning. So far, user feedback has been generally positive, and participation is further increasing. Looking to the future, we believe that the platform's model capabilities will continue to improve, and we will also deepen the integration of "XiaoTuan" within the Meituan App. Our goal is to use the new generation of AI technology to make Meituan the preferred platform for meeting various local life service needs of users. To this end, we will continue to strengthen the platform's AI search capabilities and task execution capabilities, aiming to upgrade Meituan into an industry-leading AI-driven application and an important AI entry point in the local life service field. (Continuously updating...) ### Related Stocks - [MEITUAN-W (03690.HK)](https://longbridge.com/en/quote/03690.HK.md) ## Related News & Research - [Alset AI Announces Trading on the OTCQB Venture Market in the United States and 10-for-1 Share Consolidation to Enhance Shareholder Value](https://longbridge.com/en/news/281011224.md) - [Nebius to Construct 310 MW AI Factory in Finland | NBIS Stock News](https://longbridge.com/en/news/281145974.md) - [Ericsson's Vonage to Provide Broot.ai With Voice APIs](https://longbridge.com/en/news/281508301.md) - [The Bull Case For Honeywell International (HON) Could Change Following New U.S. Defense And AI Deals](https://longbridge.com/en/news/281342974.md) - [EQT invests in space-based AI data centers - Di Digital](https://longbridge.com/en/news/281153394.md)